Jianyi
Wang‡
a,
Menghui
Chen‡
b,
Weiwei
Qin
*ac and
Meng
Zhou
c
aFaculty of Geosciences and Environmental Engineering, Southwest Jiaotong University, Chengdu 611756, China. E-mail: weiwei.qin@swjtu.edu.cn
bInstitute for Sustainable Energy/College of Sciences, Shanghai University, Shanghai, 200444, China
cChemical and Materials Engineering Department, New Mexico State University, Las Cruces, NM 88003, USA
First published on 2nd July 2022
Lithium–sulfur batteries (LSB) offer a high energy density in energy storage systems in the long run, and are of much lower cost than commercially available lithium-ion batteries. The electrochemistry performance of LSB can be improved greatly through the electro-catalytic mechanism, especially when it is applied on the separator, facing towards the cathode. However, due to the limitations of current technology (vacuum filtration and the slurry coat on polypropylene), high performance LSB's development is stunted. In this paper, an ultra-thin coating and quick methods were investigated to improve the performance of LSB by a synergy between a reduced graphene oxide (RGO) loaded S-catalyst (O-vacancy-enriched –BaTiO3) and glass fiber separator, which has excellent fire retardant and high temperature performance (cycled more than 100 times at 60 °C), and can greatly prolong the service life of the LSB. The following theories and experiments show that the original separator modified design produces the electrochemical and safety performance of the LSB, which is significant for both theoretical research and vast application prospects.
The main components of Celgard series separators are polypropylene (PP) and polyethylene (PE), which hardly limit the diffusion of polysulfide. The reason is that, because of its large pore size, polysulfide can pass through easily by electric field drive and concentration difference between the cathode and anode areas.4,5 At present, there are great challenges in the interface structure design and surface modification of lithium metal anodes. The separator modifications include special mechanisms to inhibit lithium dendrite formation and polysulfide diffusion, which is expected to promote the development and practical application of LSB.6,7 In addition, most research focuses on the modification of PP separators, which can hinder the diffusion of polysulfide and improve the performance of LSB (Scheme 1a).8 Functional design of the modified separator is one of the feasible and simple methods to improve the performance of LSB. So far, many new functional separators have shown great potential in practical battery applications. By introducing functional materials coatings on the separator, including polymer materials, carbon-based materials,9–12 inorganic oxides,13,14 catalytic polar metal compounds15 and metal/covalent organic frameworks,16–20 the specific capacity, cycle stability, multiplier performance, Coulomb efficiency and safety of LSB can be significantly improved. There are two main preparation technologies: one is to use the vacuum suction filtration and barrier layer which were placed on the surface of the PP separator, but this method takes a long time, especially for nanomaterials; the other one would be make a paste of barrier with binder and coat on the PP separator, yet the thickness of the layer cannot be controlled. Therefore, developing a new type of preparation technology which can integrate the advantages of composite separators and get rid of their disadvantages is of critical importance.
In addition, the thermal safety of LSB is particularly important due to the use of lithium metal anodes and sulfur cathodes.21 At present, thermal safety research in the field of LSB is in urgent need of further research. Most of the existing modified coating separators are based on PP separator modifications, so their thermal stability is poor.22 Therefore, choosing glass fiber membranes for sodium ion batteries presents a price advantage23 and their good heat resistance as separator has great advantages. In addition, it is essential to hinder the diffusion of polysulfide so it is trapped in cathode area and reactivate the deposition of deactivated sulfur-containing substances, which enhance the utilization active material. The modified coating separator applied to LSB not only needs to have properties of electrostatic repulsion, steric resistance, physical and chemical adsorption and other ways to control the polysulfide, but also needs to combine with the conductive medium to reduce the interface resistance between the cathode and separator, so as to effectively improve the utilization rate of active substances in the cathode area. Recently, a novel composite cathode material (LiFePO4/graphite) for dual-ion batteries has been proposed. The composite cathode combines LFP, with high capacity and high stability, and graphite, with high conductivity and high voltage, showing excellent electrochemical performance.24 In addition, carbon nanotubes are combined with transition metal selenides. Carbon nanotubes have good conductivity, forming a network structure in the composite electrode, and enhance the specific capacitance of the charge path with good conductivity.25 By far the most successful example of carbon-incorporation-enhanced battery performance is graphene.26 The typical short-tubular antimony trithiophosphate SbPS4 composite with graphene oxide (SbPS4/GO) exhibits a high theoretical specific capacity according to the dual mechanism of transformation alloying. Inserting a unique short tube into the composite can promote the transport of Na ions, alleviate the huge volume change during charge and discharge, and improve the chemical performance.27 If graphene was applied in LSB, the charge transfer resistance would be reduced, the utilization rate of the electrode active material would be improved, and the energy density of battery would be increased.28
In this paper, a reduced graphene oxide supported oxygen vacancy enriched BaTiO3 (OV-BaTiO3@RGO) was selected as the separator coating because the pure carbon material is not fully encapsulated in the separator and has cracking and fracture sites, preventing conventional carbon materials from capturing polysulfide (Scheme 1b). In this regard, the strategies of ferroelectric barium titanate (BaTiO3) as an additive29 and PP separator coating30 have been proved. However, it is difficult to suppress the polysulfide filling within the structure of RGO and each particle (Scheme 1c). In general, the long-term cycling performance of LSB remains to be improved. The mixture of OV-BaTiO3@RGO with ethanol is uniformly loaded on the surface of the glass fiber diaphragm by a suction filtration method to improve the conductivity and utilization of S. The oxygen vacancy of OV-BaTiO3 with ferroelectricity and electrocatalytic activity as a separator coating achieves a synergy with fixed soluble polysulfide and a catalytic effect, which can be the further adsorption of polysulfide dissolved in the electrolyte, and provides the active material sulfur with electrochemical reactivity to increase the specific capacity (Scheme 1d). Reduced capacity decays by 0.09% during the cycle (1 A g−1). An excellent reversible capacity of 720 mA h g−1 was obtained after 200 cycles, even with a high sulfur content (90%) and load (10 mg cm−2) and a high area capability of 7 mA h cm−2was maintained. In order to reveal the mechanism of the electrochemical reaction, a series of electrochemical experiments were carried out. In situ XRD results show that there is a strong interaction between the charge and discharge products of S during cycles, and the CV results of the polysulfide symmetric battery also confirmed the catalytic effect on Li2S. Combined with theoretical calculations (DFT), we found that the dynamic change process has a strong effect on polysulfide with OV-BaTiO3. In addition, the modified glass fiber separator can not only achieve the ideal flame retardant effect, but also maintain the excellent electrochemical performance of LSB.
It was found using EPR that the OV-BaTiO3 from alkali metal reducing agent assisted fabrication keeps high amounts of oxygen vacancies, which increases ion transport in Li–S battery systems. The introduction of oxygen vacancies can change the electron band structure of the polysulfide catalyst, enhance the binding with polysulfide, and accelerate the surface electron exchange for a rapid electrochemical reaction rate.31 In addition, the surface of oxygen vacancies provides highly active adsorption sites for diverse polysulfides and can increase the utilization rate of the S cathode.
The adsorption energy can be calculated as follows: E = E(slab + Li2Sx) − E (slab) − E (Li2Sx).
In situ XRD technology is an important method in current energy storage research analysis.33 It cannot only eliminate the impact of external factors on the electrode material (compare with the ex situ XRD), but also improve the authenticity and reliability of the data when monitoring of the process of the electrochemical reaction in real-time for characterizing the structure and composition change. Thus, the whole reaction of the system can be analysed and processed more clearly, and its intrinsic reaction mechanism can be revealed. Therefore, the introduction of in situ XRD characterization technology can improve the understanding of the reaction features of the catalytic conversion of liquid polysulfide (shuttle factor) to solid polysulfide, which characterizes the charge–discharge process of OV-BaTiO3 as the electrocatalyst layer of LSB and to explore its reaction mechanism, as shown in Fig. 2a. Through the characterization, the formation of Li2S is confirmed. However, with the standard pure cellgard separator as a reference it can be seen that S8 peaks exist over the whole cycle of the battery test, indicating that the shuttling of polysulfides is serious.34 Strengthening the chemical adsorption of the oxygen vacancies and permanent electric polarization of OV-BaTiO3 and polysulfide is the key to improving the cyclic stability of LSB. Through “chemical adsorption” between nanoparticles of OV-BaTiO3 and soluble polysulfide during the charging and discharging process, insoluble products that cannot be dissolved in the electrolyte are formed, to effectively inhibit the loss of active substances.
The cathode using the pure S electrode (without C/S composite material) combined with the OV-BaTiO3@RGO@GF separator has highest performance at the same current density. The initial capacities of LSB with OV-BaTiO3@RGO@GF separator, BaTiO3@RGO@GF separator, RGO@GF separator and GF separator at 1 A g−1 current density are 897, 849, 720 and 273 mA h g−1 (Fig. 2b), respectively. Moreover, it can be seen that the LSB with the OV-BaTiO3@RGO@GF separator cycled 200 cycles retains a capacity of 720 mA h g−1, with a low-capacity decay of 0.08% per cycle. These results indicate that OV-BaTiO3 is able to effectively inhibit the shuttle effect and reduce the polysulfide shuttle. Its unique properties of oxygen vacancy provide a driving force for polysulfide conversion and accelerate the electrochemical kinetics process. Then the mechanism of OV-BaTiO3's specific contribution to battery capacity is explored. It can be further seen from the variation of the voltage–capacity curve that the LSB with a OV-BaTiO3@RGO@GF separator has a significantly higher discharge capacity from Fig. 2c.
At the end of the charge–discharge cycles test, the battery is tested using an electrochemical impedance spectrum (EIS) and the result shows that OV-BaTiO3@RGO@GF separators have a low-impedance (Fig. S1, ESI†). These tests are a further demonstration that oxygen vacancy has outstanding abilities. In addition, the infiltration experiment further proved the adsorption effect of OV-BaTiO3 with polysulfide (Fig. 2d). The CV curve of the Li2S6 symmetric battery shows that OV-BaTiO3 electrode has good reversibility and fast redox kinetics, indicating that the inherent electrochemical activity of OV-BaTiO3 can effectively accelerate the redox kinetics of active substances. In addition, from the adsorption experiment (Fig. S2, ESI†), the color change of OV-BaTiO3@RGO after adsorbing Li2S6 was the most obvious (compared with BaTiO3@RGO and RGO), which further indicated that OV-BaTiO3@RGO had strong adsorption performance for Li2S6. The above results showed that O-vacancy-enriched BaTiO3 could effectively adsorb polysulfides under the synergistic effect of O-vacancy and ferroelectricity, and effectively inhibit the shuttle effect. The influence of catalytic conversion is more significant in the presence of OV-BaTiO3 compared with RGO (Fig. S3, ESI†). In the study of rate performance (Fig. 2e), the discharge capacities of LSB with OV-BaTiO3@RGO@GF separators at 0.5–10 A g−1 are 1184, 1048, 986, 882 and 693 mA h g−1, respectively. The physisorption and chemisorption without catalytic conversion of the BaTiO3@RGO@GF separator is only 388 mA h g−1 at 10 A g−1. These results indicate that the LSB with OV-BaTiO3@RGO@GF separator has some distinct advantages for improving the capacity attenuation caused by the shuttle effect, but it can guarantee the polysulfide does not shuttle to the anode and effectively improves the utilization rate of sulfur at ultra-high current density. Under the condition of 2 A g−1 (Fig. S4, ESI†), the LSB with the OV-BaTiO3@RGO@GF separator maintains a reversible capacity of 410.6 mA h g−1 at 400th cycle. The decay rate is as low as 0.06% per cycle (382 mA h g−1 after 600 cycles). Finally, when the surface load of S was further increased to about 10 mg cm−2 (Fig. S5, ESI†), the LSB had a considerable initial capacity (1358 mA h g−1, 13.58 mA h cm−2) at 0.05 A g−1. In addition, after 18 cycles at 0.2 A g−1, it can still maintain a reversible capacity of approximately 724 mA h g−1 (7.24 mA h cm−2), which is higher than the recommend ideal LSB area capacity.35
In the DFT examination of battery materials, ion/molecule adsorption energy is often calculated. By comparing the adsorption energy with that of the active material, the performance of electrode materials can be estimated, and then the final theoretical results and the experimental data can be compared and analyzed.36 Many LSB systems have been researched to solve the problems of the dissolution and shuttling of polysulfide with DFT calculations and to analyse the adsorption capacity of polysulfide on the substrate. Therefore, the adsorption energies between polysulfide and OV-BaTiO3 and BaTiO3 were simulated by the DFT method (Fig. 3a). The adsorption energies of OV-BaTiO3 on S8, Li2S8, Li2S6, and Li2S4 were −4.083, −5.782, −5.481 and −3.973 eV, respectively, indicating that the adsorption of OV-BaTiO3 on LiPSs was achieved by the OV–S bond, which was consistent with the expected results. Although the adsorption energy of Li2S2, and Li2S for OV-BaTiO3 is lower than that of BaTiO3; adsorption energies of BaTiO3 for Li2S2 and Li2S are −2.799 and −2.848 eV, respectively, which are much higher than that of OV-BaTiO3. There was a widely held belief that shuttle effect involves soluble polysulfide (Li2S4-8), in which solid polysulfide (Li2S1-4) can be ignored.
Further analysis of the rate-limiting step and the calculation of adsorption energy involved is below (Fig. 3b):
ΔE = E(total) − E(slab) − E (*) |
Wherein, the adsorption of Li2Sx is compared to the molecular energy of Li2Sx in vacuum:
ΔE (Li2S8) = E (Li2S8 + slab) − E (slab) − E (Li2S8) |
The S2/S of each desorption step is compared to the energy of the two Li2Sx and the desorption of S2:
Li2S4 → Li2S2, then S2 = Li2S4 − Li2S2 |
ΔE(S2) =E(S2 + slab) − E(slab) − E(Li2S4 − Li2S2) |
Li2S4 → Li2S2, then S2 = Li2S6 − Li2S4 |
ΔE(S2) = E(S2 + slab) − E(slab) − E(Li2S6 − Li2S4) |
Li2S2 → Li2S, then S= Li2S2–Li2S |
ΔE(S) = E(S + slab) − E(slab) − E (Li2S2 − Li2S) |
From here we can draw the conclusion that the rate-limiting step of a perfect surface of BaTiO3 is Li2S2 desorption and S is 3.108 eV. Instead, the rate-limiting step of oxygen vacancies of OV-BaTiO3 surface is 1.031 eV. Compared with the perfect surface, the reaction rate-limiting step is greatly reduced, so the formation of oxygen vacancies is conducive to the occurrence of reaction.
Although such design can be performed at a higher level at room temperature, we also need to think about the point of batteries' thermal safety characteristics.37 For this reason, as we mentioned earlier, it is a good idea to use a higher temperature of the LSB. The oxygen vacancy of OV-BaTiO3, conductive graphene, glass fiber skeleton of ternary composite structure, and the multicomponent integrated design can not only guarantee the coating fully maintaining its electrical conductivity, but also could effectively suppress the shuttle effect of polysulfide and ensure that the cycle stability of the battery can be maintained at 60 °C (Fig. 4a). The OV-BaTiO3@RGO@GF separator maintained good charge and discharge stability under a current density of 1 A g−1 with a high capacity of 490 mA h g−1. Further, the performance of battery using the GF separator is lower than that of any other modified separators because more polysulfides shuttle in the first few cycles. Next, the heat treatment with ultra-high temperature (280 °C) for OV-BaTiO3@RGO@GF separator was tested and achieved good specific capacity (620 mA h g−1 at 2 A g−1). Even more surprising, the battery remained safe after 200 cycles (Fig. 4b). From the images in Fig. 4b, there were no apparent changes of the OV-BaTiO3@RGO@GF separator observed after the heat treatment. On the contrary, the heat changes the OV-BaTiO3@RGO@PP separator into carbon (Fig. S6, ESI†), which proves how reliable and good the OV-BaTiO3@RGO@GF separator is. Further vertical firing tests are used to investigate the flame retardancy of the intumescent flame retardant on OV-BaTiO3@RGO@GF separator and OV-BaTiO3@RGO@PP separator (Fig. 4c and d), and the OV-BaTiO3@RGO@GF separator shows good fire retardancy. These experimental results indicate that those properties of the coatings with OV-BaTiO3 can reach the requirements of the weathering performance and fire retardancy of high performance LSB.
Footnotes |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2ma00576j |
‡ These authors contributed equally. |
This journal is © The Royal Society of Chemistry 2022 |